Thanks, both of you. I see Gforth-itc will be fine for me.
On Sat, Aug 16, 2025, at 5:58 PM, Anton Ertl wrote:
> On Sat, Aug 16, 2025 at 08:40:20AM +0100, Ethan Azariah wrote:
>>I rarely used "fast" because I read that it has reduced error checking; is
>>this still true?
>
> Yes, gforth-fast detects data stack underflows later or not at all.
> E.g., try
>
> : foo drop 0 ;
> clearstack
> foo
>
> In addition, gforth-fast does not check for integer division errors
> (division by zero or overflow on division). If the CPU catches it
> (AMD64), gforth-fast will report it, however.
>
>>(I don't need the speed in my interactive use.)
>
> Then gforth-itc should do it. If something does not work, please
> report it. I never use it myself, so I don't notice any breakage.
>
> Concerning speed: On a 2400MHz Cortex-A76 (numbers are times in
> seconds, smaller is better):
>
> sieve bubble matrix fib fft
> 0.155 0.120 0.040 0.094 0.039 gforth-fast
> 0.364 0.383 0.196 0.302 0.158 gforth
> 0.430 0.465 0.330 0.524 0.194 gforth-itc
>
> - anton
